align(4909d): every rule arm that annotates reads one bar, kb_rule_threshold at 0.70

prompt_rule, pre_tool_rule and write_path_rule now share one settings key,
RULE_FLOOR_KEY = "kb_rule_threshold", default 0.70. The reply backstop keeps
its own 0.80, because there the floor is a stop bar. Moving the shared floor
through tune_retrieval or the Settings form records a tuning event on every
surface that reads it (set_dial, via retrieval_surfaces.sharing_floor), and
the result names the surfaces it carried as `moved_with`. Settings shows one
"Standing-rule confidence threshold" control where it showed three.

Operator ruling (2026-10-05): one shared bar for the rule arms, "especially
since we're only taking the first x number of them" -- the budget limits
what is shown; the bar only decides what is worth ranking. Kept a dial.

Why 0.70, from records read rather than percentiles:
- write_path_rule near-misses at 0.7199-0.72 were all false positives
  (rule 149 against a Lidarr comment, 157 against a retention Svelte file,
  121 against an Archives restore, 164 against a config test, 131 against a
  Go package), so the old 0.72 was not hiding real hits -- but the band does
  noise control on this arm, so dropping 0.02 costs little.
- prompt_rule: the operator set 0.70 by hand on 2026-09-18, after a 0.60
  experiment admitted junk and was reverted. 0.70 is the number with a human
  decision behind it.
- pre_tool_rule lines near 0.68 this session were mostly noise; rising to
  0.70 trims them, and the band still narrows what clears.

Retired keys: kb_promptrule_threshold, kb_toolrule_threshold and
kb_rulehint_threshold are no longer read. Rows an install stored under them
stay in the settings table, inert. On this instance the effective bars move
pre_tool_rule 0.68 -> 0.70, write_path_rule 0.72 -> 0.70, prompt_rule stays
0.70; record_release_defaults writes those per-surface release rows on boot,
so the telemetry shift has an event to explain it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

src/grimoire/services/plugin_context.py

@@ -144,7 +144,7 @@ WRITEPATH_THRESHOLD_KEY = SURFACES["write_path"].floor_key
 WRITEPATH_DEFAULT_ENABLED = True
 WRITEPATH_DEFAULT_THRESHOLD = SURFACES["write_path"].floor_default
 
-# The standing-rule arm gets its own bar, for a third corpus: 0.68 above is
+# The standing-rule arms get their own bar, for a third corpus: 0.68 above is
 # calibrated on code against note PROSE, not code against RULE TEXT.
 #
 # THE STRUCTURAL ARGUMENT, which holds on any install, including one with six
@@ -155,64 +155,27 @@ WRITEPATH_DEFAULT_THRESHOLD = SURFACES["write_path"].floor_default
 #      pool always returns something, so "the best match cleared the bar"
 #      drifts from "a good match exists" toward "N things were ranked". A bar
 #      calibrated for best-of-thousands is cleared by best-of-forty as
-#      arithmetic rather than relevance. (The arms search every tier, which
-#      makes the pool larger; the bar still errs toward silence on purpose.)
+#      arithmetic rather than relevance.
 #   2. Rules are short imperative technical English — a far more HOMOGENEOUS
 #      corpus than note prose, and a more homogeneous corpus has a HIGHER
 #      noise floor than prose's 0.55-0.63.
 #
-# WHY 0.72 AND NOT A NUMBER OFF A HISTOGRAM. The exact offset between prose's
-# floor and rule-text's is not derivable in general — it depends on how an
-# install writes its rules — so the default errs deliberately toward SILENCE
-# rather than toward recall, on an asymmetry that is itself structural: this
-# hint fires on EVERY write. A missed rule is recoverable, because the rule is
-# still in Grimoire and the agent can search it. A hint that cries wolf is not:
-# it teaches the reader to skip the whole block, and the surface is lost along
-# with the true positives it would have carried.
+# ONE BAR FOR EVERY RULE ARM THAT ANNOTATES. The write path, the command arm
+# and the prompt arm read the same key (`RULE_FLOOR_KEY` in retrieval_pipeline):
+# they search the same rules, and what limits each one is its budget and band.
 #
 # TUNE IT FROM YOUR OWN INSTANCE: `retrieval_telemetry` reports
-# `rule_usage.pull_through`. Raise this if rules arrive unread; lower it if
-# rules you needed never arrived. What would RETIRE it: a cross-encoder rerank,
-# which would make a similarity bar the wrong control entirely.
-#
-# This bar is the WRITE-PATH arm's. A code payload is long and rich, which is
-# the case 0.72 suits; the command arm has its own bar below.
+# `rule_usage.pull_through` and `near_miss_samples`. Raise it if rules arrive
+# unread; lower it if rules you needed never arrived. What would RETIRE it: a
+# cross-encoder rerank, which would make a similarity bar the wrong control
+# entirely.
 RULEHINT_THRESHOLD_KEY = SURFACES["write_path_rule"].floor_key
 RULEHINT_DEFAULT_THRESHOLD = SURFACES["write_path_rule"].floor_default
 
-# THE COMMAND ARM'S OWN BAR, AND WHY IT IS NOT THE WRITE PATH'S.
-#
-# A write-path query is a code payload, long and rich, while a pre-tool query
-# is a shell command — often under a dozen words. Less text, less signal, lower
-# scores for the same relevance. At the write path's 0.72 the command arm
-# turns nearly everything away, with its refused mass piled just under the
-# line, which is the shape a bar set too high leaves.
-#
-# WHAT 0.68 IS CALIBRATED AGAINST. Replayed queries, consequential acts
-# against innocuous ones, scored against a rule corpus with symptom-bearing
-# triggers:
-#
-#   0.7571  git push origin dev                 consequential
-#   0.7245  cd ...; git fetch; git add -A       consequential
-#   0.7193  git pull --rebase origin dev        consequential
-#   0.6850  docker compose up -d                consequential
-#   ----------------------------------------  0.68
-#   0.6735  wc -l src/*.py && date              innocuous
-#   0.6544  grep -rn useState src/              innocuous
-#   0.6099  sed -n '120,160p' package.json      innocuous
-#   0.6056  ls -la && cat README.md             innocuous
-#
-# At 0.72 three of the four consequential acts would retrieve nothing, though
-# the right rules ranked first for each.
-#
-# THE SEPARATION IS NARROW (about 0.01), and that is a caveat, not a result.
-# Eight probes set a direction; they do not settle a number.
-# `near_miss_samples` on real traffic is what settles it, and this is the bar
-# to re-read first.
-#
-# WHY A LOWER BAR IS SAFE HERE. The band does noise control downstream: a marginal hit that
-# clears the bar still has to score within `_RULEHINT_BAND` of the top to be
-# rendered. The bar's job shrank, so the bar can.
+# THE COMMAND ARM READS THE SAME BAR. A shell command is short, so it scores
+# lower than a code payload for the same relevance and the shared bar admits
+# less here. The band does noise control downstream: a hit that clears the bar
+# still has to score within `_RULEHINT_BAND` of the top to be rendered.
 #
 # The innocuous scores near the line are largely CROSS-PROJECT BLEED rather
 # than bad ranking: another project's shell-command rule matching a shell
@@ -478,17 +441,11 @@ def _autoinject_query(prompt: str, context: str) -> str:
 # The operator's message is the only query that exists before one is composed,
 # and this arm is what runs against it.
 PROMPTRULE_THRESHOLD_KEY = SURFACES["prompt_rule"].floor_key
-# INHERITED FROM THE ACT ARMS. 0.72 is tuned against
-# code and shell commands. An operator's prose is a different query shape
-# against the same documents, and nothing yet says the two distributions line
-# up — triggers are written in the vocabulary of the MOMENT, which for most
-# rules is act vocabulary, so prose may well score lower across the board.
-#
-# Starting at the act arms' number anyway is deliberate: it is the only value
-# with evidence behind it, and guessing lower would put an unmeasured bar in
-# front of a corpus that binds. Every call is logged under `prompt_rule`, so
-# real traffic settles it — read `near_miss_samples` before moving this, not
-# the percentile alone.
+# THE SAME BAR AS THE ACT ARMS. An operator's prose is a different query shape
+# against the same documents, and triggers are written in the vocabulary of
+# the moment, which for most rules is act vocabulary. Every call is logged
+# under `prompt_rule`, so read `near_miss_samples` before moving the shared
+# bar, not the percentile alone.
 PROMPTRULE_DEFAULT_THRESHOLD = SURFACES["prompt_rule"].floor_default
 
 # ANCHORED ON THIS HOOK'S BUDGET rather than the act arms'. This arm fires
@@ -1171,9 +1128,8 @@ async def get_writepath_config(user_id: int) -> dict:
     independent pairs, resolved from the registry.
 
     Each floor's reason is recorded where its default is: WRITEPATH (code
-    embeddings sit on a much higher similarity floor than prose), RULEHINT (a
-    third corpus again), TOOLRULE (a shell command is a different query shape
-    from a code payload and scores lower for the same relevance).
+    embeddings sit on a much higher similarity floor than prose) and RULEHINT
+    (a third corpus again, whose one bar both rule arms read).
 
     THE BUDGET IS SPLIT TOO. "How many titles at once" does not mean the same
     thing on both surfaces: this arm fires before every Write and Edit while